A Longitudinal Study of Maternal Depressive Symptoms, Negative Expectations and Perceptions of Child Problems

A Longitudinal Study of Maternal Depressive Symptoms, Negative Expectations and Perceptions of Child Problems Ilona Luoma, MD Pälvi Kaukonen, MD Mirjami Mäntymaa, MD Kaija Puura, MD, PhD Tuula Tamminen, MD, PhD Department of Child Psychiatry, Tampere University Hospital, University of Tampere, Medical School Raili Salmelin MSc, PhD School of Public Health, Tampere University Hospital, Research Unit, University of Tampere ABSTRACT: The aim of this longitudinal study was to examine the associations between maternal depressive symptoms and perceptions of children s problems. One hundred and nineteen mother child dyads were followed from the third trimester of pregnancy for almost 10 years. Depressive symptoms and background factors of the mothers and the anticipated/perceived problems of their firstborn were assessed prenatally, postnatally, and when the child was 4 5 years and 8 9 years old. The simultaneous and long-term associations between maternal depressive symptoms and child s problems were examined. Maternal prenatal depressive symptoms, the continuity of negative expectations to postnatal problem perceptions, and high problem level at 4 5 years of child s age predicted high problem level in 8 9-year-olds. KEY WORDS: maternal depression; Child Behavior Checklist (CBCL); Edinburgh Postnatal Depression Scale (EPDS); Neonatal Perception Inventory (NPI). Many studies have shown that children of affectively ill parents do worse than children of parents without psychiatric problems, although many of the children also show resilience. 1 Postnatal depression, in particular, has been the focus of many recent studies Address Correspondence to Ilona Luoma, MD, University of Tampere, Medical School, FIN-33014 University of Tampere; e-mail: ilona.luoma@uta.fi. Child Psychiatry and Human Development, Vol. 35(1), Fall 2004 2004 Springer Science+Business Media, Inc. 37

38 Child Psychiatry and Human Development and the adverse impact of postnatal depression on children is well documented. 2,3 Evidence from follow-up studies of postnatal depression shows long-term cognitive and academic consequences for children, at least under conditions where depression is associated with other risk factors. 4,5 The harmful impact of maternal depression on a child begins prenatally. Newborn infants born to mothers who were depressed during pregnancy have been found to be less active, have poorer muscle tone, and to be more irritable and irregular in their sleeping rhythm than newborns of healthy mothers. 6 It has been documented that depressed mothers prenatal norepinephrine and dopamine levels predict the corresponding hormone levels of the newborn and suggested that maternal depression during pregnancy may have an early biochemical influence on neonatal outcome. 7 In our earlier study, the presence of mother s depressive symptoms during pregnancy predicted child s behavioral difficulties at 8 9 years of age. 8 Allen, Lewinsohn and Seeley also reported that maternal (retrospectively recalled) emotional problems during the pregnancy predicted major depression and disruptive behavior disorder of their children in adolescence. 9 Attachment theory 10 provides a different framework for understanding and explaining the relationship between maternal depression and child development. Mother s postnatal depression may contribute to difficulties in the formation of an affective bond with the infant, 11 and depression during pregnancy may have the same impact, since the birth of attachment in parents occurs during the pregnancy. 12 Mothers have representations of their infants long before conception. 13 Mothers prenatal representations of their infants are documented to show considerable stability and be associated with infants attachment classifications at the age of 12 months. 14 Mother s representation of attachment during pregnancy is also predictive of attachment status later 15, and prenatal attachment predicts early mother infant interaction. 16 Maternal depression affects many domains of mother infant interaction and increases the risk for insecure attachment. 17 As insecurity of attachment increases and the attachment pattern becomes less adaptive, the risk of psychopathology increases. 18 A third possible mechanism of transmission is the mother child interaction, which is the context in which the mother establishes her perceptions concerning the baby. Mother s emotional impairment affects her perceptions of her child. 19 Mothers who are depressed or stressed have been shown to interact with their children more negatively, with more commands and criticism, which in turn may

Ilona Luoma et al. 39 elicit further problematic behavior on the part of the children. 20 Depressed mothers stimulate their infants less through talk and play, express more negative feelings toward them and are less sensitive to their cues than non-depressed mothers. 2,21 It has been shown that objectivity and subjectivity in maternal perceptions are not easily distinguishable, and that maternal reports reflect both mother s and child s well-being. 22 It also has been shown that early perceptions of child s problems tend to have continuity and predict subsequent child socioemotional problems. 23 25 Aims The first aim of this study was to investigate whether maternal depressive symptoms during pregnancy and during the postnatal period affect maternal expectations and perceptions of the baby. We assumed that maternal depressive status is associated with more negative maternal expectations during pregnancy and increased problem perceptions of the child postnatally and later, reflecting maternal negative representations of the child. The second aim was to examine the issue of continuity and discontinuity in maternal problem perception concerning the child from perinatal stages to 4 5 years and 8 9 years of child s age. The hypothesis was that mother s negative expectations during pregnancy and negative early perceptions concerning the baby would be associated with high level of subsequent problem perceptions. Study Design Method This study is a part of a prospective follow-up study that started in Tampere, Finland, in 1989. 26 At the first stage, maternal depressive symptoms in a community sample of healthy first-time mothers were screened by means of questionnaires during late pregnancy (T1) and three times postnatally (T2 = after delivery, T3 = 2 months and T4 = 6 months postnatally). Maternal (prenatal) expectations and (postnatal) perceptions of the baby were assessed at the same time points. Of the postnatal measurements T3 was selected for analysis in this report due to a known incidence peak of postnatal depression at 2 3 months 27 and the first biobehavioral shift in infant development at the same age. 28 At the second and third stages of the follow-up study in 1994 1995 (T5) and in 1997 1998 (T6) maternal depressive symptoms were screened again and maternal reports

40 Child Psychiatry and Human Development of the firstborn s emotional and behavioral status were gathered by means of questionnaires. Sample The flow chart of the longitudinal study is shown in Figure 1. The sample was collected from all the maternity health clinics in Tampere, Finland, during a 6-month period in 1989 1990. The original sample of 349 mothers who agreed to participate in the study represented healthy Finnish firsttime mothers. Only 10% of the target population declined to participate. A group of 279 mothers (Sample A) was included in a more intensive study design. 26 Seventy-eight of the 279 mothers dropped out of the study during the first six months after delivery. The largest group of drop-outs (69 mothers) consisted of mothers who did not receive the questionnaires of the first postnatal survey in the obstetrics ward in the hospital during the summer. The drop-out groups at the perinatal stage, particularly the first and largest group of drop-outs, were analyzed carefully using prenatal data. They did not differ significantly from the respondent group by sociodemographic characteristics or by maternal prenatal level of depressive symptoms. 26 A group of 70 mothers (Sample B), who also participated in the study, was not included in the more intensive part of the longitudinal study due to time schedule of Study stage 1, and was also excluded at T5, but included again at T6 (Figure 1). In Sample A, between T4 and T5 one mother withdrew from the study due to serious illness of the child, and thus at T5 and T6 questionnaires were sent to 200 mothers of the perinatal study stages (the follow-up sample). The questionnaires were completed at T5 by 158 recipients (79%) and at T6 by 142 of the 200 mother child pairs (71%). Sample A consisted of those mother child pairs for whom complete mother child data was available from stages T1, T3, T5 and T6 (n = 119), which is 34% of the original sample and 60% of the follow-up sample. Due to relatively small sample size and the possibility of selective attrition, complementary analyses were also accomplished for the 165 mother child dyads on whom complete data was available from stages T1, T3, and T6 (Sample A + B). The size of Sample A + B is 47% of the original sample and 61% of the sample of 270 mothers included in the follow-up at T6. Thus, the data from T5 is not included in these supplementary analyses. To assess the possibility of selective attrition at the later stages, the drop-outs of both Sample A and Sample A + B were compared with respondents with respect to parental educational level, employment, marital status, previous psychiatric problems, health status and maternal prenatal level of depressive symptoms on the basis of data gathered at T1. Drop-outs did not differ statistically significantly from the study samples with respect to any of these background factors. Both Samples A and A + B included more mother daughter dyads than mother son dyads: 42% of the children were boys in Sample A and 43% in Sample A + B, whereas 64% were boys in the original sample (p = 0.001) and 66% (p = 0.001) among the followup sample drop-outs. The mothers in Sample A were younger than those among the follow-up sample drop-outs (maternal age below 24 years during the initial data collection 32% in Sample A and 11% among the follow-up

Ilona Luoma et al. 41 Figure 1. Flow Chart of the Study.

42 Child Psychiatry and Human Development sample drop-outs, 25 29 years 46% vs. 59%, 30 years or older 22% vs. 31% respectively; p = 0.002). Measures The depressive symptoms of the mothers at each of the time points were screened using the Edinburgh Postnatal Depression Scale (EPDS). The EPDS is a self-report questionnaire originally designed for screening depression among women during the postpartum period, but it has been found also to have satisfactory validity among non-postnatal women. 29,30 When completing the EPDS the mothers are asked to choose from the options those that best describe their feelings during the previous seven days. The scale consists of 10 items scored on a four-step scale from 0 to 3, the sum score of items thus ranging from 0 (no depressive symptoms) to 30 (high level of depressive symptoms). With a cutpoint of 12/13 (also used in the present study), the sensitivity was 64% and specificity 96% for postnatal depression. 26 For non-postnatal major depression, sensitivity of 88% and specificity of 80% have been reported. 30 The Neonatal Perception Inventory (NPI) is a questionnaire designed for the assessment of the mother s perceptions of her baby. 23 The mother is asked to assess the crying, vomiting, feeding, bowel movements, sleeping, and predictability of the behavior of her own baby (Your Baby) and of an average baby (Average Baby). The six items of both Your Baby and Average Baby sections are scored on a five-step scale from 1 (no concerns) to 5 (very many concerns). The sum score varies between 6 and 30 for both sections. The NPI score is obtained by subtracting the Your Baby sum score from the Average Baby sum score. According to Broussard and Hartner, the NPI score is categorized into better than average (positive) if the mother reports fewer problems in her baby than in an average baby and not better than average (negative), when the mother reports as many or even more problems in her baby compared with an average baby. 23 The proportions of positive and negative NPI scores from time points T1 (pregnancy) and T3 (2 months) are reported. Because the combined predictive ability of successive NPI reports has been reported to be greater than an NPI report at a single time point, 23 a combined measure of NPI reports was also formed by categorizing the NPI reports into those being negative at both time points and those including a positive NPI report at one or both of the time points. The mothers completed the Child Behavior Checklist (CBCL) questionnaires 31 when the children were 4 5 years old (T5) and 8 9 years old (T6). The CBCL is an internationally used instrument designed to record children s competencies and problems as reported by their parents. The CBCL problem scale includes 118 items, each of which is scored on a three-step scale from 0 (item not true) to 2 (item very true or often true). The total problem sum scores were converted into normalized T scores, which were used as child outcome variables at both of the later stages. The cutpoint of 59/60 was used to separate problematic and non-problematic children. Sociodemographic and health data on the mothers was gathered at T1, T5 and T6 by questionnaires designed for this study.

Ilona Luoma et al. 43 Statistical Methods For descriptive purposes frequencies were used. Associations of pairs of categorized variables were examined by cross-tabulations, together with two-tailed Pearson chi square test, or Fisher s exact test, as appropriate. P-values less than 0.05 are interpreted as significant, and p-values less than 0.10 are reported as nearly significant. The simultaneous effects of potential predictors of problems at school age were analyzed by logistic regression. The results of the regression analysis are presented as odds ratios (ORs) and their 95% confidence intervals (95% CIs). The analyses were accomplished with SPSS for Windows version 9.0. Sample Characteristics Results All of the mothers were Finnish and represented by sociodemographic characteristics Finnish first-time mothers. 32 The characteristics of the samples are presented in Table 1. Table 1 Background Characteristics of the Mothers in the Sample A and Sample A + B at 8 9 years of child s age (T6) Sample A Sample A + B (n = 119 ) (%) (n = 165 ) (%) Marital status Married or cohabiting 91 87 Single 9 13 SES of the family a Upper 85 84 Lower 15 16 Gender of the child Girl 58 57 Boy 42 43 Age of the mother 34 or younger 48 47 35 or older 52 53 a Socioeconomic status (SES) of the family was determined by the occupation of the main breadwinner; for single mothers families the SES of the family was determined by the occupation of the mother. The upper category consisted of white collar workers and enterpreneurs, the lower category of blue collar workers, students and others.

44 Child Psychiatry and Human Development Maternal Depressive Screening Status and Child Problems In Sample A, 9% of the mothers had significant depressive symptoms prenatally, 6% postnatally, 8% at 4 5 years and 6% at the latest stage. In Sample A + B the respective percentages were 12% prenatally, 9% postnatally, and 6% at the latest stage. In Sample A, 42% of the babies had negative NPI report at T1 and 23% at T3. Twenty-one percent of children had high problem level according to the CBCL at T5 and 16% at T6. The associations tested, along with their p-values between mother s depressive status at different time points and her expectations of the baby and perceptions of the child are shown in Figure 2. Maternal concurrent depressive symptoms were not statistically significantly associated with negative expectations of the baby or negative perceptions of the child. However, at school-age (T6) the simultaneous association was nearly significant. The association between maternal prenatal depressive symptoms and child problems at T6 was statistically significant. The association between maternal depressive symptoms at T5 and high level of child problems at T6 was nearly significant (Figure 2). In Sample A+B the proportions of negative NPI reports were 41% at T1 and 25% at T3. The percentage of children with high scores in the CBCL at T6 was 16% as in Sample A. None of the associations between maternal concurrent depressive symptom level and the level of problem perception was statistically significant even though at T6 the difference in percentages of high scoring children between the groups seemed high (15% vs. 33%, p = 0.166). Maternal prenatal depressive symptoms were associated with high scores in the CBCL at T6 (p = 0.004). Continuity and Discontinuity of Problem Perceptions The frequencies of high scores in the CBCL reports and their associations with positive and negative NPI groups are presented in Table 2. In Sample A, although there was a tendency towards higher levels of problem perceptions at T5 among the high-scorers in the NPI, none of the associations was statistically significant at p < 0.05 level. If the NPI remained negative at both time points, T1 and T3, the risk of high problem level at T6 was statistically significantly higher compared to children with positive NPI report at either or both time points (Table 2). The presence of high problem level at T5 was significantly associated with high problem level at T6 as well.

Ilona Luoma et al. 45 Figure 2. Associations Between the Presence of Maternal Depressive Symptoms and the Level of Child Problems at Successive Time Points. The Nonsignificant Associations (p 0.1) are Presented as Broken Lines. In Sample A + B, negative NPI report at T1 was also statistically almost significantly associated with a high score in the CBCL at T6 (Table 2).

46 Child Psychiatry and Human Development Table 2 Maternal expectations and perceptions in the NPI (T1 = third trimester of pregnancy, T3 = two months postnatally) and the frequencies (%) of high problem reports according to the CBCL at 4 5 years (T5) and at 8 9 years (T6) High CBCL Total Problems Sample A at T5 Sample A at T6 Sample A + B at T6 (n = 117 a ) (n = 119 ) (n = 165 ) Maternal Perceptions % p % p % p T1 (NPI) 0.353 0.202 0.052 Positive 18 12 11 Negative 27 22 24 T3 (NPI) 0.291 0.371 0.174 Positive 19 14 14 Negative 31 22 24 Combined 0.175 0.047 0.018 Perceptions T1 and T3 (NPI) Positive at T1 19 13 14 and/or at T3 Negative at 36 36 37 both T1 and T3 T5 (CBCL Total 0.029 problems) Low 12 High 32 a n at T5 is lower due to missing data in the CBCL. Factors Predicting Behavioral/Emotional Problems at 8 9 Years of Age Factors associated significantly (p < 0.05) or nearly significantly (p < 0.10) with high problem level at T6 according to previous analyses (negative NPI at both time points, T1 and T3, maternal depressive symptoms at T1, T5 and T6, and high CBCL problem level at T5) were included in a logistic regression analysis together with child s gender, maternal age, marital status and family socioeconomic status (presented in Table 2). In Sample A, high problem level in the CBCL at T6 was predicted by prenatal maternal depressive symptoms (OR 6.3, 95% CI 1.6 25.1), high CBCL problem level at T5 (OR 3.2, 95% CI 1.0 9.7) and negative NPI reports both preand postnatally (OR 3.6, 95% CI 1.0 13.3).

Ilona Luoma et al. 47 When the analysis was repeated in Sample A + B (data from T5 and maternal depressive symptoms at T6 excluded; persistence of negative NPI reports from T1 to T3, maternal depressive symptoms and negative NPI report at T1 as well as background factors included), the factors remaining in the model were prenatal maternal depressive symptoms (OR 5.4, 95% CI 1.9 15.7) and negative NPI reports both pre- and postnatally (OR 4.2, 95% CI 1.4 12.4). Discussion The main objective in this study was to concentrate particularly on mother s perceptions of her child, and the associations between perceptions and depressive symptoms over time. In the light of these results, the relationships between mother s depressive symptoms and the perceptions of problems in her child are variable over time. Long-term associations and continuities in problem perceptions are discernible. The lack of simultaneous associations between maternal depressive symptoms and child problems at prenatal and postnatal stages was unexpected. At the later stages the finding may be due to the small sample size: the number of mothers with significant depressive symptoms was small and therefore statistical significance was not reached, although the perceptions of the symptomatic mothers were somewhat more negative than those of non-symptomatic mothers. However, prenatally and postnatally the percentages of negative NPI reports were essentially the same among mothers with and without depressive symptoms, and not even a trend towards greater percentages of negative reports among the children of mothers with depressive symptoms was detected. This suggests that the negativity of the early expectations and perceptions, at least when measured by the NPI, is a separate phenomenon from maternal depressive symptomatology. Early maternal expectations and perceptions seem to have an independent role in the occurrence of subsequent problem perceptions. The findings may be understandable from the perspective of developmental psychopathology. It may be that the effect of maternal depression on a child cannot be seen simultaneously at the early stage but later when a child with lowered adaptive capacity is facing cumulative risk factors during development. According to the same theoretical background it is possible to explain why the prenatal depressive symptomatology of the mother predicted child problems years later.

48 Child Psychiatry and Human Development The persistence of negativity from prenatal to postnatal stage predicted child s high problem level at the age of 8 9 years, suggesting that an early negative trajectory of perceptions has continuity over the following years. The continuity of certain aspects in maternal perceptions has earlier been reported in other studies than those of Broussard and Hartner, 23 Bates, Freeland and Lunsbury have reported a significant correlation between high NPI scores at 1 month and fussy-difficult temperamental rating in the Infant Characteristic Questionnaire at 4 6 months of age. 33 It has also been suggested that parents may shape their child s developing temperament according to their perceptions. 34 Mother s perceptions and the child s actual characteristics are both important for child development. Mother s negative expectations during pregnancy may have been based at least in some cases on actual difficulties during the pregnancy (problems in fetal growth, for example). In these cases the negative expectations during pregnancy and negative perceptions at two months may both have a common foundation: the actual developmental problems of the child apart from difficulties of the mother or in the interaction. These child characteristics may have the persistence, which is seen in the maternal reports. The findings should be interpreted with caution because the sample size was relatively small and the proportion of drop-outs was considerable, when the number of mother child dyads included in the follow-up is considered. However, the proportion of respondents and the attrition rate are comparable with other longitudinal studies with follow-up time as long as in this study. 4 The main findings were similar for both samples analyzed in the present study, which supports the stability and generalizability of the results. The background characteristics of the follow-up group likewise did not differ significantly from the drop-out group, except for child gender and maternal age. The fact that more mother son dyads dropped out of the study may be due to boys on average having more behavioral problems compared with girls. 35 Mothers having trouble with their child may have been less willing to participate because of child problems or other reasons associated with the situation. Non-responding mothers may also have had more concurrent depressive symptoms, although this issue could not be measured on the basis of data available. If the drop-out groups included more problematic children and symptomatic mothers, their participation in the study would probably have strengthened the association between simultaneous maternal depressive symptoms and problem perceptions concerning the child.

Ilona Luoma et al. 49 The problems emerging through categorization should be mentioned. The dichotomization of mothers into low- and high-scorers by a cutpoint in the screening questionnaire limits the consideration of dimensionality in depression. Green 36 argues that valuable data is wasted by dichotomizing and suggests that the EPDS could be used as a continuous measure of mothers emotional well-being pre- and postnatally. The majority of mothers span a continuum of depressive symptoms, and depressive symptom levels below the (rather high) cutpoint in EPDS may not be insignificant for a child. For practical purposes, however, the established cutpoint was used in this study and the issue of dimensionality must be returned to. The number of mothers scoring high in the EPDS was somewhat lower compared, for example, with the results of a recent cohort study conducted in the UK, reporting 13.5% of mothers scoring high on the EPDS at 32 weeks of pregnancy and 9.1% at 8 weeks postpartum. 37 These lower proportions among the respondents in this study, however, are not likely to be due to selective attrition of depressed mothers during the follow-up, since the proportions of high-scoring mothers among the respondents of the follow-up did not differ from the proportions in the whole sample at the first study stage. 26 The lower proportion of high-scorers may be due to cultural factors: an international study exploring levels of postpartum depressive symptoms showed considerable variation in EPDS scores between countries, with lowest mean scores in Sweden, Australia and Finland. 38 Low scores in Finland may be due to the well-functioning social welfare and health service system (nearly all mothers attending maternity care and well-baby clinics), and a relatively good standard of living in a community with fairly homogeneous ethnic and social structure. It could be argued that one limitation is the use of questionnaires to define caseness : clinical interviews and observations made for smaller subsamples of mothers and children were not considered in this report. On the other hand, Hammen, Adrian, Gordon, Burge, Jaenicke & Hiroto 39 demonstrated that, compared with a mother s diagnosed affective disorder, self-reported depressive symptoms of mothers were related to a greater extent to children s difficulties (reported not only by mothers but also by teachers), and concluded that the relations between maternal psychiatric status and child disorder are a complex issue. Many studies have examined the possible negative distortion of mother s perceptions due to depression. The results have been contradictory, 40 showing depression-related distortion of perception, 19,41 or increased accuracy, 42 or mixed results reporting bias being present

50 Child Psychiatry and Human Development for boys but not girls, 43 and being significant but not accounting for the majority of variance. 44 Probably both arguments are true: depressed mothers tend to see their children more negatively and their children tend to have more problems. Maternal bias or the objective validity of maternal reports, however, was not the question to be addressed in the present study. We focused specifically on the relation between a mother s perceptions of her child and symptoms of depression experienced by mothers, and their relations over time. The measures of child problems used in the present study are also measures of maternal problem perception, not diagnostic instruments. Furthermore, it should be emphasized that no gold standard exists to determine whether there are real or objective problems in the child or not. The problem reports of any informants may be biased, 40,45 and not all child problems really present at home are detected at school or in clinical interviews or observations. The context- and relationship-specificity of children s symptoms has long been acknowledged. 46 Generally, in both clinical and research settings the mother is still the primary source of information concerning infants and young children, and is usually the person who knows her child best. Summary The results of this study suggest that the harmful impact of maternal depression on a child begins during the pregnancy. The relative importance of the prenatal period highlights the significance of the early interplay between biology and psychology, which forms the basis of attachment and later interaction. Professionals in maternal and child health care should pay particular attention to the persistence of maternal negativity concerning the baby from pregnancy to the postnatal period because this is a risk factor for the continuity of problem perceptions. Acknowledgements The study was supported by grants from the Finnish Child Psychiatric Research Foundation, the Foundation of Paediatric Research, the Medical Research Fund of Tampere University Hospital and the Odd Fellow Association.

Ilona Luoma et al. 51 References 1. Beardslee WJ, Versage EM, Gladstone TRG: Children of affectively ill parents: a review of the past 10 years. J Am Acad Child Adolesc Psychiat 37: 1134 1141, 1998. 2. Murray L, Fiori-Cowley A, Hooper R, Cooper P: The impact of postnatal depression and associated adversity on early mother infant interactions and later infant outcome. Child Developmt 67: 2512 2526, 1996. 3. Murray L, Sinclair D, Cooper P, Ducournau P, Turner P, Stein A: The socioemotional development of 5-year-old children of postnatally depressed mothers. J Child Psychol Psychiat 40: 1259 1271, 1999. 4. Hay DF, Pawlby S, Sharp D, Asten P, Mills A, Kumar R: Intellectual problems shown by 11-year-old children whose mothers had postnatal depression. J Child Psychol Psychiat 42: 871 889, 2001. 5. Kurstjens S, Wolke D: Effects of maternal depression on cognitive development of children over the first 7 years of life. J Child Psychol Psychiat 42: 623 636, 2001. 6. Field T: Infants of depressed mothers. Inf Beh Developmt 18: 1 13, 1995. 7. Lundy BL, Jones NA, Field T, Nearing G, Davalos M, Pietro PA, Schanberg S, Kuhn C: Prenatal depression effects on neonates. Infant Beh Developmt 22: 119 129, 1999. 8. Luoma I, Tamminen T, Kaukonen P, Laippala P, Puura K, Salmelin R, Almqvist F: Longitudinal study of maternal depressive symptoms and child well-being. J Am Acad Child Adolesc Psychiat 40: 1367 1374, 2001. 9. Allen NB, Lewinsohn PM, Seeley JR: Prenatal and perinatal influences on risk for psychopathology in childhood and adolescence. Dev Psychopathol 10: 513 529, 1998. 10. Bowlby J: Attachment. London: Penguin Books, 1969. 11. Kumar R: Anybody s child : severe disorders of mother-to-infant bonding. Br J Psychiat 171: 175 181, 1997. 12. Brazelton TB, Cramer BG: The Earliest Relationship. Parents, Infants and the Drama of Early Attachment. London: Karnac Books, 1991. 13. Stern D: The Motherhood Constellation. A Unified View of Parent Infant Psychotherapy. New York: Basic Books, 1995. 14. Benoit D, Parker KCH, Zeanah CH: Mothers representations of their infants assessed prenatally: stability and association with infants attachment classifications. J Child Psychol Psychiat 38: 307 313, 1997. 15. Fonagy P, Steele H, Steele M: Maternal representations of attachment during pregnancy predict the organization of infant mother attachment at one year of age. Child Developmt 62: 891 905, 1991. 16. Siddiqui A, Hägglöf B: Does maternal prenatal attachment predict postnatal mother infant interaction? Early Hum Developmt 59: 13 25, 2000. 17. Teti DM, Gelfand DM, Messinger DS, Isabella R: Maternal depression and the quality of early attachment: an examination of infants, preschoolers, and their mothers. Dev Psychol 31: 364 376, 1995. 18. Boris NW, Zeanah CH: Disturbances and disorders of attachment in infancy: an overview. Infant Mental Health J 20: 1 9, 1999. 19. Najman JM, Williams GM, Nikles J, Spence S, Bor W, O Callaghan M, Le Brocque R, Andersen MJ, Shuttlewood GJ: Bias influencing maternal reports of child behaviour and emotional state. Soc Psychiatry Psychiatr Epidemiol 36: 186 194, 2001. 20. Webster-Stratton C: Mothers and fathers perceptions of child deviance: roles of parent and child behaviors and parent adjustment. J Consult Clin Psychol 56: 909 915, 1988.

52 Child Psychiatry and Human Development 21. Murray L, Cooper PJ: The role of infant and maternal factors in postpartum depression, mother infant interactions, and infant outcome. In: Postpartum Depression and Child Development, eds. Murray L, Cooper PJ. New York: The Guilford Press, 1997, pp. 111 135. 22. Foreman DM, Henshaw C: Objectivity and subjectivity in postnatally depressed mothers perceptions of their infants. Child Psychiat Hum Developmt 32: 263 275, 2002. 23. Broussard ER, Hartner MSS: Further considerations regarding maternal perception of the first born. In Exceptional Infant 2. Studies in Abnormalities, ed. Hellmuth J. New York: Butterworths, 1971, pp. 432 449. 24. Rende RD: Longitudinal relations between temperament traits and behavioral syndromes in middle childhood. J Am Acad Child Adolesc Psychiat 32: 287 290, 1993. 25. Guerin DW, Gottfried AW: Temperamental consequences of infant difficultness. Infant Beh Developmt 17: 413 421, 1994. 26. Tamminen T: Postnatal Depression, Breastfeeding and Mother infant Interaction. Doctoral dissertation (in Finnish with a summary in English), Vol. 304, Tampere: Acta Universitatis Tamperensis ser A, 1990. 27. Kumar R, Robson KM: A prospective study of emotional disorders in childbearing women. Br J Psychiat 144: 35 47, 1984. 28. Zeanah CH, Boris NW, Larrieu JA: Infant development and developmental risk: a review of the past 10 years. J Am Acad Child Adolesc Psychiat 36: 165 178, 1997. 29. Cox JL, Holden JM, Sagovsky R: Detection of postnatal depression. Development of the 10-item Edinburgh Postnatal Depression Scale. Br J Psychiat 150: 782 786, 1987. 30. Cox JL, Chapman G, Murray D, Jones P: Validation of the Edinburgh Postnatal Depression Scale (EPDS) in non-postnatal women. J Affective Dis 39: 185 189, 1996. 31. Achenbach TM: Manual for the Child Behavior Checklist/4-18 and 1991 Profile. Burlington VT: University of Vermont Department of Psychiatry, 1991. 32. Statistics Finland (1997). 33. Bates JE, Freeland CAB, Lounsbury ML: Measurement of infant difficultness. Child Developmt 50: 794 803, 1979. 34. Pauli-Pott U, Mertesacker B, Bade U, Haverkock A, Beckmann D: Parental perceptions and infant temperament development. Inf Beh Developmt 26: 27 48, 2003. 35. Kraemer S: The fragile male. Br Med J 321: 1609 1612, 2000. 36. Green JM: Postnatal depression or perinatal dysphoria? Findings from a longitudinal community-based study using the Edinburgh Postnatal Depression Scale. J Reprod Infant Psychol 16: 143 155, 1998. 37. Evans J, Heron J, Francomb H, Oke S, Golding J, on behalf of the Avon Longitudinal Study of Parents and Children Study Team: Cohort study of depressed mood during pregnancy and after childbirth. Br Med J 323: 257 260, 2001. 38. Affonso DD, De AK, Horowitz JA, Mayberry LJ: An international study exploring levels of postpartum depressive symptomatology. J Psychosom Res 49: 207 216, 2000. 39. Hammen C, Adrian C, Gordon D, Burge D, Jaenicke C, Hiroto D: Children of depressed mothers: maternal strain and symptom predictors of dysfunction. J Abnorm Psychol 96: 190 198, 1987. 40. Richters JE: Depressed mothers as informants about their children: a critical review of the evidence for distortion. Psychol Bull 112: 485 499, 1992. 41. Najman JM, Williams GM, Nikles J, Spence S, Bor W, O Callaghan M, Le Brocque R, Andersen MJ: Mother s mental illness and child behavior problems: cause effect association or observation bias? J Am Acad Child Adolesc Psychiat 39: 592 602, 2000.

Ilona Luoma et al. 53 42. Conrad M, Hammen C: Role of maternal depression in perceptions of child maladjustment. J Cons Clin Psychol 57: 663 667, 1989. 43. Boyle MH, Pickles A: Maternal depressive symptoms and ratings of emotional disorder symptoms in children and adolescents. J Child Psychol Psychiat 38: 981 992, 1997. 44. Youngstrom E, Izard C, Ackerman B: Dysphoria-related bias in maternal ratings of children. J Cons Clin Psychol 67: 905 916, 1999. 45. Treutler CM, Epkins CC: Are discrepancies among child, mother, and father reports on children s behavior related to parents psychological symptoms and aspects of parent-child relationships? J Abnorm Child Psychol 31: 13 27, 2003. 46. Achenbach TM, McConaughy SH, Howell CT: Child/adolescent behavioral and emotional problems: implications of cross-informant correlations for situational specificity. Psychol Bull 101: 213 232, 1987.